Windows is huge compared to a typical Linux server setup -- Server 2003 takes up 20 times as much disk space as Ubuntu's server offering, and on the desktop, it's still a factor of three or four. On one hand, a lot of that is going to be help files, images, GUIs, and so forth; on the other, there's just going to be a lot more executable code that might be running.
This isn't an excuse for Windows to have exploits, but it's probably a large portion of the cause.
Or report them to a large portion of the affected people. Let them contact the principal. More steps between you and the people whose heads will roll is better. And a principal dealing with 200 angry teachers, most of whom already have a scapegoat (the principal), won't have time to find another, with any luck.
Even if he was looking for something to hack, he didn't do any damage. Instead, he performed a public service. Punishing a person for something he maybe was wanting to do is just stupid.
On the other hand, if he didn't phrase his message carefully, it could have been taken as a threat. If he said something along the lines of "Please use a more secure password on $SERVER -- I guessed it easily", then it's hard to sympathize with the administration. If he said "I accessed your server and now have the social security numbers for every faculty member", then it's much more ambiguous, and I'd expect the student to be investigated. Just investigated, not arrested.
In the past, I've experienced, say, Excel loading in a third the time of OOCalc. I've also experienced OOCalc loading much faster than Excel, and these experiences are much more recent.
If there's a speed difference between the two, it'll probably change within the next year or two. I wouldn't use that as a basis for choosing between them.
You want to protect your kids online. Fine. Who is going to rape your kids via the Internet?
Be careful about who your kids see in person, if you're worried. But they're going to run in to everything from Tubgirl to 4chan to scientology on the internet; you can't filter it out. Make sure they're well grounded in reality, and they should be able to handle it.
That's a thin shield. I'm much happier with a citizen army than any alternative, but I'd be even happier if I could register a militia and keep AKs and so forth in an appropriately secured facility, as a private citizen. I acknowledge that this would put me under greater scrutiny than normal citizens.
I think that National Guard troops would be willing to disregard orders to harm innocent citizens, but they could be misled without much trouble. Also, I think they would have trouble opening up their armories to random citizens except in the most extreme circumstances. Even if the US government turned entirely fascist and authoritarian, I doubt that would suffice, unless it were accomplished by violence.
The question was specifically for an opt-out filtering system. The new plan is for a filtering system that does not allow anyone to opt out. Therefore it is much less likely to gain support and much more likely to gain opposition. For the lazy, the opt-out system had 13.3% support and 73.8% opposition.
And the way in which the FCC will handle that is by fining the manufacturers of the GSM chip. If software can do such things to the network, you're dealing with a software radio, which has much different requirements for certification and distribution than a hardware radio.
The thing is that Android allows for installing programs from -- hear and be astonished! -- other sources than Google itself, unlike Apple. Without any extra or undue inconvenience.
Android has that capability, yes.
And the manufacturers can disable that capability.
If I buy an Android phone through Verizon, for example, it'll be locked so I can only use applications that I buy through approved vendors -- Google or Verizon. Verizon is evil, but it isn't likely to offer anything interesting anyway. Google might not be evil yet, unlike Apple, but I don't want to bet on its remaining good. That's why I use a Freerunner.
You can, in fact, develop Fennec for mobile OSX. Any version released through the iphone app store could not include Javascript capabilities, though -- people could write their applications in js and release them through other channels in that case.
A default install of the latest official image from Openmoko gets you a reasonable, basic phone. It doesn't take any special effort to use it. The default install doesn't have all the features of the latest Windows Mobile phone, I admit. But there's a fair bit of software that you can add.
The only sucky part is the keyboard -- it's fine with a stylus, or a stick, or a mechanical pencil, but not so great with fingers. That's mainly a side effect of the narrowness of the screen. I'm hoping the next revision has a larger screen.
Oh, dear. I guess I'll have to stop using my Freerunner, then.
GSM chips are so loosely connected with the rest of the phone that this scenario isn't likely. And the networks would be idiots to assume that all GSM chips would behave nicely, or know what the relative load on each tower is, or what issues there are with the tower...
You can't depend on your clients, but even your open clients aren't powerful.
There are two weak spots when using unbreakably strong encryption: - Poor choice in passwords - Rubber-hose cryptography
Both are probably going to be sufficiently useful that algorithm strength won't be a significant issue. The exception to that is if people start encrypting large volumes of innocuous data.
Someone who can analyse the source code and determine that it is just parsing responses without any reason to think it's understanding.
If it is simply parsing responses, then that will likely be insufficient to convince you, even without disclosing the source code.
Though you do bring up the point of understanding. What is that? I'd say that it is the ability to alter your internal state based on new information, and then behave differently in the future based on that information. Most of the chat bots I've seen competing don't even have a memory of what they heard three lines ago.
I believe I have a better record on instant messages than one defect per fifteen minutes, if you ignore punctuation. It's not difficult to achieve a low defect rate; it just takes a couple seconds to ensure that you are typing correctly.
It's helpful that both firefox and pidgin have builtin spellcheck functionality, but even when using pidgin on windows (I haven't set up a dictionary for it there), I keep a low defect rate.
Note also that explicit parallelism isn't going to get anywhere near optimal, either.
Or rather, there's a PTAS for optimal multithreaded code, and the cost for getting code better than an automated parallelism scheme will quickly approach the cost of creating that automated parallelism scheme.
Because 90% of developers understand Java, and maybe 10% understand SmallTalk. TIOBE lists SmallTalk as #36 in popularity with 0.123% market share and Java as #1 with over 20%.
Granted, TIOBE is based on search engine results, which aren't a perfect indicator of usage, but they are probably accurate to the order of magnitude.
I prefer Linux, and I want Linux to be supported by my hardware vendor. Therefore I am willing to pay a small additional fee to indicate my preference to the vendor. This will encourage them to continue supporting Linux in the future.
Not to mention that their interest would in keeping you subscribing to the service they're currently offering, with a minimum of support calls, and therefore in improving their product's feature set, usability, and stability.
Slashdot Mirror
Windows is huge compared to a typical Linux server setup -- Server 2003 takes up 20 times as much disk space as Ubuntu's server offering, and on the desktop, it's still a factor of three or four. On one hand, a lot of that is going to be help files, images, GUIs, and so forth; on the other, there's just going to be a lot more executable code that might be running.
This isn't an excuse for Windows to have exploits, but it's probably a large portion of the cause.
Or report them to a large portion of the affected people. Let them contact the principal. More steps between you and the people whose heads will roll is better. And a principal dealing with 200 angry teachers, most of whom already have a scapegoat (the principal), won't have time to find another, with any luck.
The kid was trying to be anonymous -- just because of asshats like this. How is that blackmail?
Even if he was looking for something to hack, he didn't do any damage. Instead, he performed a public service. Punishing a person for something he maybe was wanting to do is just stupid.
On the other hand, if he didn't phrase his message carefully, it could have been taken as a threat. If he said something along the lines of "Please use a more secure password on $SERVER -- I guessed it easily", then it's hard to sympathize with the administration. If he said "I accessed your server and now have the social security numbers for every faculty member", then it's much more ambiguous, and I'd expect the student to be investigated. Just investigated, not arrested.
In the past, I've experienced, say, Excel loading in a third the time of OOCalc. I've also experienced OOCalc loading much faster than Excel, and these experiences are much more recent.
If there's a speed difference between the two, it'll probably change within the next year or two. I wouldn't use that as a basis for choosing between them.
You want to protect your kids online. Fine. Who is going to rape your kids via the Internet?
Be careful about who your kids see in person, if you're worried. But they're going to run in to everything from Tubgirl to 4chan to scientology on the internet; you can't filter it out. Make sure they're well grounded in reality, and they should be able to handle it.
I, for one, applaud the fine men and women who are doing their duty to protect us from the ravages of Canadia.
That's a thin shield. I'm much happier with a citizen army than any alternative, but I'd be even happier if I could register a militia and keep AKs and so forth in an appropriately secured facility, as a private citizen. I acknowledge that this would put me under greater scrutiny than normal citizens.
I think that National Guard troops would be willing to disregard orders to harm innocent citizens, but they could be misled without much trouble. Also, I think they would have trouble opening up their armories to random citizens except in the most extreme circumstances. Even if the US government turned entirely fascist and authoritarian, I doubt that would suffice, unless it were accomplished by violence.
http://whirlpool.net.au/survey/2007/
The question was specifically for an opt-out filtering system. The new plan is for a filtering system that does not allow anyone to opt out. Therefore it is much less likely to gain support and much more likely to gain opposition. For the lazy, the opt-out system had 13.3% support and 73.8% opposition.
This is what happens when you put too much power in the hands of just one hundred men!
And the way in which the FCC will handle that is by fining the manufacturers of the GSM chip. If software can do such things to the network, you're dealing with a software radio, which has much different requirements for certification and distribution than a hardware radio.
The thing is that Android allows for installing programs from -- hear and be astonished! -- other sources than Google itself, unlike Apple. Without any extra or undue inconvenience.
Android has that capability, yes.
And the manufacturers can disable that capability.
How much do you want to bet they won't?
If I buy an Android phone through Verizon, for example, it'll be locked so I can only use applications that I buy through approved vendors -- Google or Verizon. Verizon is evil, but it isn't likely to offer anything interesting anyway. Google might not be evil yet, unlike Apple, but I don't want to bet on its remaining good. That's why I use a Freerunner.
You can, in fact, develop Fennec for mobile OSX. Any version released through the iphone app store could not include Javascript capabilities, though -- people could write their applications in js and release them through other channels in that case.
There's the bling module for e17 that supports transparency and other effects. I haven't found a use for it, but it exists.
A default install of the latest official image from Openmoko gets you a reasonable, basic phone. It doesn't take any special effort to use it. The default install doesn't have all the features of the latest Windows Mobile phone, I admit. But there's a fair bit of software that you can add.
The only sucky part is the keyboard -- it's fine with a stylus, or a stick, or a mechanical pencil, but not so great with fingers. That's mainly a side effect of the narrowness of the screen. I'm hoping the next revision has a larger screen.
Oh, dear. I guess I'll have to stop using my Freerunner, then.
GSM chips are so loosely connected with the rest of the phone that this scenario isn't likely. And the networks would be idiots to assume that all GSM chips would behave nicely, or know what the relative load on each tower is, or what issues there are with the tower...
You can't depend on your clients, but even your open clients aren't powerful.
There are two weak spots when using unbreakably strong encryption:
- Poor choice in passwords
- Rubber-hose cryptography
Both are probably going to be sufficiently useful that algorithm strength won't be a significant issue. The exception to that is if people start encrypting large volumes of innocuous data.
But atleast it smelt nice.
The iron I got from it was exquisite.
Someone who can analyse the source code and determine that it is just parsing responses without any reason to think it's understanding.
If it is simply parsing responses, then that will likely be insufficient to convince you, even without disclosing the source code.
Though you do bring up the point of understanding. What is that? I'd say that it is the ability to alter your internal state based on new information, and then behave differently in the future based on that information. Most of the chat bots I've seen competing don't even have a memory of what they heard three lines ago.
I believe I have a better record on instant messages than one defect per fifteen minutes, if you ignore punctuation. It's not difficult to achieve a low defect rate; it just takes a couple seconds to ensure that you are typing correctly.
It's helpful that both firefox and pidgin have builtin spellcheck functionality, but even when using pidgin on windows (I haven't set up a dictionary for it there), I keep a low defect rate.
Note also that explicit parallelism isn't going to get anywhere near optimal, either.
Or rather, there's a PTAS for optimal multithreaded code, and the cost for getting code better than an automated parallelism scheme will quickly approach the cost of creating that automated parallelism scheme.
Because 90% of developers understand Java, and maybe 10% understand SmallTalk. TIOBE lists SmallTalk as #36 in popularity with 0.123% market share and Java as #1 with over 20%.
Granted, TIOBE is based on search engine results, which aren't a perfect indicator of usage, but they are probably accurate to the order of magnitude.
I prefer Linux, and I want Linux to be supported by my hardware vendor. Therefore I am willing to pay a small additional fee to indicate my preference to the vendor. This will encourage them to continue supporting Linux in the future.
Not to mention that their interest would in keeping you subscribing to the service they're currently offering, with a minimum of support calls, and therefore in improving their product's feature set, usability, and stability.